Artificial Intelligence at CSL

How we’re integrating AI into our global biotech business

Story
artificial intelligence illustration

When CSL assembled last year at its Data Science Summit in Bern, Switzerland, the company brought together participants from every aspect of its operations, from Finance to Pharmacometrics.

With data science and artificial intelligence playing an ever-increasing role, Global Head of Digital Transformation and Execution Systems Karen Etchberger posed a question relevant to all: “How do we move from where we are today to a future where we can bring this ambition to life?”

John Thompson, CSL’s Global Head of Advanced Analytics and Artificial Intelligence, is helping to lead that effort. Thompson has been on the forefront of artificial intelligence and its use in business for more than 30 years. He has helped build analytic systems for giants of global industry like Coca-Cola, Anheuser-Busch and Dell. Now, he’s doing the same for CSL.

“This company has grown tremendously and has been making smart moves along the way,” Thompson said. “Now it’s really starting to see the value of data and analytics.”

Developing a Data Science Framework

CSL is taking a two-pronged approach that includes a Center of Excellence and a Community of Practices on advanced analytics and AI, Thompson said. Data scientists are part of both groups and working on numerous projects throughout CSL. The setup ensures each project is the right one to address individual team needs while benefitting CSL as a whole.

“There’s a lot do,” Thompson said. “My days go from dawn to dusk every day and I feel as energized as I did when I started in the morning. It’s an exciting time to be here.”

Artificial intelligence and machine learning can be used to comb through vast sets of data to find outliers or similarities that can illuminate understanding of any number of scenarios. Like other global industries, Thompson said the company will be using artificial intelligence to improve supply chain efficiency and to comply with regulatory and legal requirements.

Making AI Work for Patients

But CSL also wants to use those powerful engines to solve the burdens faced by patients with rare and serious diseases. A major goal in CSL Behring’s AI push is to shorten the amount of time between the onset of symptoms for a patient and an accurate diagnosis. In Pharmacovigilance – the front lines of patient safety - robotics process automation can speed the flow of information and improve operational efficiency, said Richard Wolf, Executive Director, Global Clinical Safety and Pharmacovigilance.

Wolf says his team is also working with others across industry to find areas where AI and natural language processing can be utilized to ensure crucial information is readily surfaced, One day, he believes it could even serve to predict risks associated with certain medications.

“We do think there is a place for artificial intelligence in our work,” Wolf said. “And we’re taking a careful and thoughtful approach toward implementing it.”

Analytics can also help CSL uncover important medical insights from vast amounts of “real-world data,” such as physician notes in a patient’s chart. Real-world data are obtained outside of randomized controlled trials and generated during routine clinical practice. Prior to advanced analytics and AI, this information existed but it was difficult to gather and analyze.

Haley Kaplowitz, Executive Director & Global Head of Safety Sciences, is leading an organization-wide effort to employ analytics as a key tool for gaining real-world evidence to be used in decision-making across the product life cycle.

Both real-world data and real-world evidence are playing an increasing role in healthcare decisions, particularly from regulatory authorities and payers. They may also help predict patient groups at increased risk of adverse events and demonstrate product effectiveness and differentiation in the marketplace, Kaplowitz said.

“The industry is under increasing pressure to provide evidence and demonstrate the value of products to multiple stakeholders, particularly in actual clinical practice,” she added. “Real-world evidence is inherent to reach this goal and increasingly crucial to ensure patient access and commercial success.”